3D building model reconstruction from a single image

Huan Chang, Fuan Tsai

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

This study developed effective algorithms to reconstruct 3D building models from a single image. The developed algorithms extract feature lines first from photographs or paintings with perspective views. Vanishing points are calculated and feature points are determined using extracted feature lines. Coordinates of the extracted feature points are then computed based on vanishing points using single-view metrology. Several types of image were used in this study to test the performance of the developed algorithms, including computer simulated paintings, video frame cuts, close-range and overlook photos, real paintings. The results demonstrated indicate that the developed algorithms can successfully extract 3D information and reconstruct 3D building models effectively from single-view images.

Original languageEnglish
Title of host publication30th Asian Conference on Remote Sensing 2009, ACRS 2009
Pages1055-1060
Number of pages6
StatePublished - 2009
Event30th Asian Conference on Remote Sensing 2009, ACRS 2009 - Beijing, China
Duration: 18 Oct 200923 Oct 2009

Publication series

Name30th Asian Conference on Remote Sensing 2009, ACRS 2009
Volume2

Conference

Conference30th Asian Conference on Remote Sensing 2009, ACRS 2009
Country/TerritoryChina
CityBeijing
Period18/10/0923/10/09

Keywords

  • 3D reconstruction
  • Building model
  • Single view reconstruction
  • Vanishing point

Fingerprint

Dive into the research topics of '3D building model reconstruction from a single image'. Together they form a unique fingerprint.

Cite this